This invention relates generally to laser diodes and more particularly concerns an electronic simulation in which thermal effects are modeled and used for correcting laser diode output.
A single beam laser diode assembly has a single diode and usually, in a scanning system, the diode is driven by a train of image pixel information. The pixel information is used to drive the diode and therefore stimulate laser flux emission where there is a white pixel in a write white system. In a write white system, a laser is turned on to create white space on a page. Intensity of the light beam is directly proportional to the output power of the laser. In order to keep the output power of the diode constant, the temperature of the diode should be kept at a constant level. However, due to the structure of the laser diode assembly, as the pixels change, the temperature of the diode fluctuates, which in turn causes the output power of the diode and the intensity of the light beam to fluctuate.
A multiple beam diode assembly has at least two diodes in close proximity on a common substrate. Each diode is driven by a separate train of image pixel information. Again, as the pixels change, the temperature of each diode fluctuates. However, in a multiple diode system, the changing temperature of a diode also causes a temperature fluctuation in adjacent diodes. The temperature fluctuations of the adjacent diodes cause the output power and the intensity of the light beams in those adjacent diodes also to fluctuate.
A tri-level system may use one or more diodes with at least one diode operating at full on, full off, and partially on. One example of an application using a single diode tri-level system is the printing of black and white documents with a highlight color. Tri-level systems suffer from the same heating effects both in the full on and the partially on modes of the laser.
In a printing system fluctuation in the intensity of light beams causes fluctuation in the size of a printed pixel. As the intensity of the light beam decreases, so does the pixel size.
In FIG. 1, the intensity variation in a diode over time is shown due to heating and cooling effects. To illustrate the effect, a first laser beam 10 is left turned on while a second laser beam 12 from an adjacent laser is cycled from a full on position to a full off position using a step function. While the second laser beam 12 is on, heating effects on the first beam 10 cause the intensity to drift downward and finally stabilize at a lower value. The change in intensity is the drift d.sub.c. A similar curve is produced from self-heating effects in the second laser 12 when it is turned on. The self-heating effects are seen in the falling time constant t.sub.fs. The rising time constant t.sub.rc or the falling time constant t.sub.fc when compared to falling time constant t.sub.fs, produced from self-heating, is larger. When the laser beam 12 has settled after self-heating effects, the difference in the output intensity is the drift d.sub.s. When the drift d.sub.c is compared to the drift d.sub.s, the drift d.sub.c is smaller than the drift d.sub.s.
When the second laser beam 12 is turned off the heat dissipates. As the heat dissipates, the intensity of the first beam 10 drifts upward and stabilizes at a higher value. The amount the intensity changes is the intensity drift d.sub.c. The amount of time needed for the intensity to drift and stabilize is a rising time constant t.sub.rc. When measured practically, rising time constant t.sub.rc and falling time constant t.sub.fc are nearly the same. This is important in designing a simulator since both the rising time constant t.sub.rc and falling time constant t.sub.fc can then be adequately modeled using a single circuit for both. A more accurate but more complicated circuit could be built to model the rising time constant t.sub.rc and the falling time constant t.sub.fc independently of each other.
Large drifts are visible in printed pages. In FIG. 2, an enlargement of a half tone pattern that illustrates the problem is shown. The pattern consists of six scan lines s1-s6 of alternating light and dark areas. For simplicity, it will be assumed that a single diode assembly in a write white system was used to scan each of the six scan lines s1-s6 sequentially starting with scan line s1 and progressing through scan line s6. However, similar problems and effects occur in write black systems, multiple diode assemblies and tri-level systems. FIG. 2 shows the half tone pattern when it is correctly printed with no heating effects. Scan lines s1 and s5 are all black. Scan lines s2, s3, and s6 are alternating blocks of black and white. Scan line s4 is all white. In a write white system, the laser is turned on to create white space and turned off to create black space. Rectangle R1 is the first white rectangle printed on scan line s2.
FIG. 3 is the pattern shown in FIG. 2 when it is printed with a single laser diode experiencing self-heating effects. When scanning scan line s1 a laser remains off the entire time since the entire line is black. When scanning scan line s2, the laser starts scanning in the off position but shortly turns on when it enters rectangle R1. When the laser turns on, the spot intensity is at its peak and gradually diminishes and stabilizes as shown in FIG. 1. The resulting change in spot size from large to small will create sloping edges E on rectangle R1. Every time the laser turns on, the same sloping edges E will be produced until the intensity and spot size stabilize.
Similar effects can be seen in multiple laser assemblies, tri-level systems and write black systems. In multiple laser assemblies, the deviations are caused from heating effects from adjacent lasers as well as the writing laser. In tri-level systems heating effects occur at both the full on and the intermediate on states.
Minimizing heating effects or otherwise compensating for or correcting for them would result in more accurate printed pages with improved image quality.
One method for compensating for laser drift would be to use a direct, real time feedback system. However, direct feedback systems have the disadvantages of being very expensive to construct, requiring extremely fast components to effectively calculate and provide a real time corrective signal, and requiring additional light paths in the printing system. Therefore, some other method must be found. The present invention uses the idea that laser drift can be modeled accurately enough to generate an appropriate correction signal.
When the pattern dependent laser drive signal is presented to both the laser diode and its model then the laser drift can be computed using the model. The correction of the laser drift can then be implemented and presented to the model to observe the correction effects. The important requirement is to make a model which faithfully models the physics of the actual heating effects of the laser.
A compensation method based on a simulation of the laser heating effects has the advantages of being inexpensive arid reliable while requiring no additional light paths and being easy to install and adapt to many different printing systems.
Accordingly, it is the primary aim of the invention to provide a method for compensating for a variety of thermally induced effects. Further advantages of the invention will become apparent as the following description proceeds.